Advertisement

The use of hemostatic spray as an adjunct to conventional hemostatic measures in high-risk nonvariceal upper GI bleeding (with video)

Published:April 20, 2016DOI:https://doi.org/10.1016/j.gie.2016.04.016

      Background

      Endoscopic management of nonvariceal upper GI bleed (NVUGIB) can be challenging. Hemospray is a novel endoscopic hemostatic agent for NVUGIB. Its efficacy in attaining hemostasis in NVUGIB is promising, particularly with respect to technically difficult lesions. However, most of the currently available data are focused on its application as monotherapy. The aim of this study was to evaluate its efficacy as a second agent to adrenaline, or as an addition to the combination of adrenaline with either clips or a thermal device in NVUGIB.

      Methods

      Consecutive patients with Forrest 1a and 1b ulcer treated with hemostatic spray as an adjunct to conventional endoscopic hemostatic measures between July 2013 and June 2015 were included in this retrospective analysis. The endpoints were initial hemostasis, 7-day rebleeding, 30-day rebleeding, all-cause, and GI-related 30-day mortality.

      Results

      A total of 20 patients (median age, 75 years, 50% men, 60% Forrest 1a ulcer) were treated with hemostatic spray as a second agent to adrenaline, or as an adjunct to the combination of adrenaline with either clips or a thermal device. Hemostatic spray was used as a second agent to adrenaline in 40% and as a third agent to combined dual therapy in 60%. Initial hemostasis was attained in 95% with an overall rebleeding rate at 7 days of 16%. There was no difference between the 7-day and 30-day rebleeding rates. The combination of hemostatic spray and adrenaline resulted in 100% initial hemostasis and 25% 7-day rebleeding. Similarly, initial hemostasis was achieved in 92% with a 9% rebleeding rate when hemostatic spray was used as the third agent to 2 of the conventional measures. All-cause mortality was 15% with 1 GI-related death (3%).

      Conclusions

      In our single-center retrospective analysis, hemostatic spray appears promising as an adjunct to conventional methods for NVUGIB, although prospective controlled trials are needed to confirm this.

      Abbreviations:

      AH (adrenaline plus Hemospray), DH (dual therapy plus Hemospray), IQR (interquartile range), NHS (National Health Service), NVUGIB (nonvariceal upper GI bleed)
      Mortality associated with acute upper GI bleeding remains high at 10%.
      • Rollhauser C.
      • Fleischer D.E.
      Current status of endoscopic therapy for ulcer bleeding.
      • Hearnshaw S.A.
      • Logan R.F.
      • Lowe D.
      • et al.
      Acute upper gastrointestinal bleeding in the UK: patient characteristics, diagnoses and outcomes in the 2007 UK audit.
      This figure has not changed appreciably since the 1950s, in part due to an aging population. In a recent UK-wide audit, 1 in 4 patients with acute upper GI bleeding were greater than 80 years of age.
      • Hearnshaw S.A.
      • Logan R.F.
      • Lowe D.
      • et al.
      Acute upper gastrointestinal bleeding in the UK: patient characteristics, diagnoses and outcomes in the 2007 UK audit.
      Because mortality from upper GI bleeding increases steeply with age, peptic ulcer bleeding will continue to pose a significant therapeutic challenge.
      • Rockall T.A.
      • Logan R.F.
      • Devlin H.B.
      • et al.
      Incidence of and mortality from acute upper gastrointestinal haemorrhage in the United Kingdom. Steering Committee and members of the National Audit of Acute Upper Gastrointestinal Haemorrhage.
      Recent advances in endoscopic techniques have enabled effective hemostatic interventions to be deployed. The combination of adrenaline with a mechanical method (clips) or with a thermal device is the preferred strategy to achieve hemostasis.
      • Marmo R.
      • Rotondano G.
      • Piscopo R.
      • et al.
      Dual therapy versus monotherapy in the endoscopic treatment of high-risk bleeding ulcers: a meta-analysis of controlled trials.
      • Sung J.J.
      • Tsoi K.K.
      • Lai L.H.
      • et al.
      Endoscopic clipping versus injection and thermo-coagulation in the treatment of non-variceal upper gastrointestinal bleeding: a meta-analysis.
      • Gralnek I.M.
      • Dumonceau J.M.
      • Kuipers E.J.
      • et al.
      Diagnosis and management of nonvariceal upper gastrointestinal hemorrhage: European Society of Gastrointestinal Endoscopy (ESGE) Guideline.
      Indeed, monotherapy with adrenaline has largely been superseded by this approach due to an unacceptably high rate of rebleeding.
      • Vergara M.
      • Bennett C.
      • Calvet X.
      • et al.
      Epinephrine injection versus epinephrine injection and a second endoscopic method in high-risk bleeding ulcers.
      However, in 10% to 20% of patients, rebleeding occurs despite initial hemostasis with combined therapy.
      • Rollhauser C.
      • Fleischer D.E.
      Current status of endoscopic therapy for ulcer bleeding.
      Hemostatic sprays are a relatively new addition to the established hemostatic agents. Hemospray (Cook Medical, Winston-Salem, NC) is the first to be licensed in Europe, but it is not as yet U.S. Food and Drug Administration approved. It is a proprietary inorganic hemostatic powder approved for use in nonvariceal upper GI bleed (NVUGIB). The powder is propelled through a carbon dioxide pressurized catheter fed through the working channel of an endoscope and sprayed at a distance of 1 to 2 cm from the bleeding site until complete coating of the lesion and an obliteration of fresh blood is achieved. When put in contact with moisture (eg, blood or tissue) in the GI tract, the powder becomes cohesive and adhesive.
      • Barkun A.N.
      • Moosavi S.
      • Martel M.
      Topical hemostatic agents: a systematic review with particular emphasis on endoscopic application in GI bleeding.
      Its hemostatic mechanisms are 2-fold. First, it exerts a tamponade effect by forming a confluent mechanical barrier at the bleeding site. Second, it promotes thrombus formation by enhancing platelet aggregation and increasing the concentration of clotting factors.
      • Giday S.A.
      Preliminary data on the nanopowder hemostatic agent TC-325 to control gastrointestinal bleeding.
      In addition, it has been shown to shorten coagulation time in vitro.
      • Holster I.L.
      • van Beusekom H.M.
      • Kuipers E.J.
      • et al.
      Effects of a hemostatic powder hemospray on coagulation and clot formation.
      The effect of hemostatic spray on hemostasis in NVUGIB cases so far has been encouraging. Initial hemostasis was achieved in 81% to 100% with a rebleeding rate of 5% to 39%.
      • Smith L.A.
      • Stanley A.J.
      • Bergman J.J.
      • et al.
      Hemospray application in nonvariceal upper gastrointestinal bleeding: results of the survey to evaluate the application of Hemospray in the luminal tract.
      • Sung J.J.
      • Luo D.
      • Wu J.C.
      • et al.
      Early clinical experience of the safety and effectiveness of Hemospray in achieving hemostasis in patients with acute peptic ulcer bleeding.
      • Masci E.
      • Arena M.
      • Morandi E.
      • et al.
      Upper gastrointestinal active bleeding ulcers: review of literature on the results of endoscopic techniques and our experience with Hemospray.
      • Yau A.H.
      • Ou G.
      • Galorport C.
      • et al.
      Safety and efficacy of Hemospray(R) in upper gastrointestinal bleeding.
      • Holster I.L.
      • Kuipers E.J.
      • Tjwa E.T.
      Hemospray in the treatment of upper gastrointestinal hemorrhage in patients on antithrombotic therapy.
      However, most of these data were focused on the use of Hemospray as monotherapy. Moreover, the study populations were heterogeneous with respect to the underlying cause. Data on the use of hemostatic spray as a second-line hemostatic agent to adrenaline or as an adjunct to combined therapy are limited. Intuitively, these strategies could offer a significant advantage; hemostatic spray may augment hemostasis achieved by standard therapy. This approach is attractive in cases in which the application of dual therapy is challenging. Adrenaline injection occasionally masks the bleeding lesion. More frequently, the position of the ulcer renders the effective application of clips or a heater probe impossible. Finally, the fibrous nature of the ulcer base or large size of a visible vessel may mean that attempting to apply clips or a heater probe would be unlikely to succeed or potentially dangerous.
      In this study, we evaluated the effect of this hemostatic spray on initial hemostasis, rebleeding rate, and mortality when used in in high-risk patients (Figure 1) with peptic ulcer bleed. Here, hemostatic spray was used as a second agent to adrenaline, or as an adjunct to the combination of adrenaline with either clips or a thermal contact device (Figure 2).
      Figure thumbnail gr2
      Figure 2Forrest 1a duodenal ulcer after a triple therapy (a combination of adrenaline, thermal device, and hemostatic spray).

      Methods

      Study population

      This retrospective analysis reviewed 30 consecutive patients who had been treated with hemostatic spray for GI bleeding between July 1, 2013, and June 30, 2015, in a large tertiary unit. One case of hemostatic spray application to lower GI bleeding, 1 case with incomplete clinical information, and 4 cases with the following underlying causes, gastroesophageal carcinoma, Mallory-Weiss tear without visible vessel, angiodysplasia, and pangastritis, were excluded. Four patients with Forrest 2a and 2b ulcers were excluded due to the small sample size.
      The final analysis included the remaining 20 cases of Forrest 1a and 1b ulcers. Of these, 1 patient had post-ampullectomy GI bleeding, but a spurting arterial vessel was identified on the index gastroscopy. Similarly, another patient who had a Mallory-Weiss tear was included in this group because a visible vessel was treated. The nature of the lesions in these 2 cases was akin to that of a peptic ulcer, and therefore they were considered as Forrest 1a and 1b, respectively. In these cases, hemostatic spray was used either as a second agent to adrenaline or as an adjunct to the combination of adrenaline with either clips or a thermal contact device.
      The outcomes were as follows: initial hemostasis, defined as adequate hemostasis confirmed on observation and to the satisfaction of endoscopist; 7-day and 30-day rebleeding, defined as (1) clinical presentation of hematemesis or melena; (2) hemodynamic instability; and (3) drop in hemoglobin ≥2 g/L, blood transfusion of ≥4 units, necessitating further investigation such as repeat endoscopy or radiologic intervention; all-cause and GI bleed-related 30-day mortality; adverse event; and equipment failure relating to the use of hemostatic spray.
      • Sung J.J.
      • Luo D.
      • Wu J.C.
      • et al.
      Early clinical experience of the safety and effectiveness of Hemospray in achieving hemostasis in patients with acute peptic ulcer bleeding.
      For comparison, we reviewed the outcomes of 20 consecutive patients with Forrest 1a and 1b ulcer who received conventional treatment before the introduction of hemostatic spray in our department between August 2010 and June 2013.
      Patients were identified using the endoscopy logbook entries. Endoscopy reports and clinical information were obtained from Unisoft (Unisoft Medical Systems, Enfield, UK) and Trakcare (Intersystem Corporation, Cambridge, Mass), respectively. Demographics, Rockall, and Blatchford scores, antiplatelet use, findings at endoscopy, endoscopic therapeutic interventions, use of proton pump inhibitor infusions, Helicobacter pylori status, and the study endpoints were recorded. All identifiable data were replaced with study identification numbers to preserve confidentiality. The use of these data was approved by the Confidentiality (Caldicott) Guardian for NHS Lothian.
      This study was considered as a retrospective review by the local committee in our institution; hence, no formal ethical approval from the NHS Research Ethics Committee was necessary.

      Clinical procedure

      The choice of treatment was at the discretion of the attending endoscopist. All patients were adequately resuscitated before the endoscopic procedure. Hemostatic spray was used to achieve complete hemostasis if there was evidence of continued oozing after conventional therapy (adrenaline with either hemostatic clips or thermal devices). Similarly, hemostatic spray was used after adrenaline injection if the lesion was inaccessible for application of hemostatic clips or thermal devices. Hemostatic spray was delivered through a 10F catheter (Cook Medical, Winston-Salem, NC) fed through the working channel of an endoscope (Olympus, Tokyo, Japan or Fujifilm, Tokyo, Japan). Adequacy of hemostasis was confirmed on observation of the treated lesion and to the satisfaction of the endoscopist. All patients received intravenous proton pump inhibitor infusion for 72 hours (Hong Kong regimen) after endoscopy.

      Statistical analysis

      Statistical analysis was performed using IBM SPSS for Macintosh version 22.0 (IBM Corporation, Armonk, NY). The Mann-Whitney U test was used to compare continuous variables between 2 populations. The Fisher exact test was used to compare categoric variables between 2 groups. Data are presented as median interquartile range (IQR) or frequency (percentage of the total study population). A 2-sided P value of less than .05 was considered to be statistically significant.

      Results

      Hemostatic spray group

      A total of 20 patients (median age, 75 years; IQR = 11 years, 50% men) were included in the final analysis. Of these, 60% (12/20) had Forrest 1a and 40% (8/20) had Forrest 1b ulcers. The median Rockall score and Blatchford score were 8 (IQR = 2) and 14 (IQR = 6), respectively. Ten (50%) patients were on antiplatelet agents at the time of presentation.
      Hemostatic spray was used as a second agent to adrenaline (AH) in 8 patients (40%) and as an addition to combined therapy (DH) in 12 patients (60%) (Table 1). The reasons for hemostatic spray application in the AH group were as follows: lesion in a difficult site, suboptimal views, and concerns regarding perforation or torrential hemorrhage, all of which had precluded effective deployment of hemostatic clips or a thermal contact device. Hemostatic spray was used as a third agent with persistent oozing or in cases where the risk of rebleeding remained significant despite conventional dual therapy. Descriptions of individual cases are provided in Supplementary Table 1, available online at www.giejournal.org.
      Table 1Demographics and clinical characteristics of patients treated with hemostatic spray
      Clinical characteristicsAdrenaline + hemostatic spray (AH) (n = 8)Dual therapy + hemostatic spray (DH) (n = 12)P value
      Age (years)7571.73
      Women37.65
      Rockall score88.75
      Blatchford score1415.51
      Use of antithrombotic therapy55.64
      Pathology.65
      Duodenal ulcer48
      Esophageal ulcer31
      Dual pathology01
      Gastric ulcer01
      Others
      Others: Mallory-Weiss with visible vessel, ampullectomy with arterial spurt.
      11
      Data presented as median (interquartile range) or number.
      Others: Mallory-Weiss with visible vessel, ampullectomy with arterial spurt.

      Initial hemostasis and rebleeding

      Initial hemostasis was achieved in 19 of 20 patients (95%). One patient (5%) who did not achieve initial hemostasis despite the use of hemostatic spray in the DH group subsequently had gastroduodenal artery embolization to achieve hemostasis. Seven-day rebleeding occurred in 16% (3/19; median day 0, IQR = 3). Rebleeding events in the AH and DH groups were 25% (2/8) and 9% (1/11), respectively Table 2. Two of the patients with rebleeding were treated with gastroduodenal artery embolization. One patient was too frail for further intervention and subsequently died. There was no rebleeding recorded beyond day 7. Thus, the 30-day rebleeding rate remained at 15%.
      Table 2Initial hemostasis, rebleeding events, and 30-day mortality stratified according to the Forrest classification in the hemostatic spray and conventional treatment groups
      Forrest classificationHemostatic spray treatment (July 2013 to August 2015)Conventional treatment (August 2010 to June 2013)Risk of rebleeding after endoscopic treatment
      • Forrest J.A.
      • Finlayson N.D.
      • Shearman D.J.
      Endoscopy in gastrointestinal bleeding.
      in the literature
      Study population, n (%)Initial hemostasis, n (%)7-day rebleeding, n (%)Mortality at 30 days, n (%)Study population, n (%)Initial hemostasis, n (%)7-day rebleeding, n (%)Mortality at 30 days, n (%)
      1a12 (60)AH5 (42)5 (100)2 (40)1 (20)A0 (0)0 (0)0 (0)0%-69%
      • Masci E.
      • Arena M.
      • Morandi E.
      • et al.
      Upper gastrointestinal active bleeding ulcers: review of literature on the results of endoscopic techniques and our experience with Hemospray.
      • Peter S.
      • Wilcox C.M.
      Modern endoscopic therapy of peptic ulcer bleeding.
      • Laine L.
      • Peterson W.L.
      Bleeding peptic ulcer.
      DH7 (58)6 (86)1 (17)0 (0)D12 (100)9 (75)1 (11); 2
      30-day rebleeding.
      (22)
      2 (22)
      1b8 (40)AH3 (38)3 (100)0 (0)0 (0)A3 (38)2 (66)2 (100)3 (100)0%-25%
      • Masci E.
      • Arena M.
      • Morandi E.
      • et al.
      Upper gastrointestinal active bleeding ulcers: review of literature on the results of endoscopic techniques and our experience with Hemospray.
      • Peter S.
      • Wilcox C.M.
      Modern endoscopic therapy of peptic ulcer bleeding.
      DH5 (62)5 (100)0 (0)2 (40)D5 (62)5 (100)1 (20)1 (20)
      AH, Adrenaline plus hemostatic spray; DH, conventional dual therapy plus hemostatic spray; A, adrenaline alone; D, conventional dual therapy (a combination of adrenaline with hemostatic clips or a thermal device).
      30-day rebleeding.

      Mortality

      All-cause mortality at 30 days in the hemostatic spray group was 15% (3/20) (Table 2). GI bleed-related mortality at 30 days was 5% (1/20). Other causes of death were septicemia and aspiration pneumonia.

      Rebleeding event and mortality according to the severity of bleed

      Forrest 1a

      Forty-two percent (5/12) of those with Forrest 1a lesions received hemostatic spray as a second line to adrenaline therapy (AH) (Table 2). In this group, initial hemostasis was achieved in 100%. However, rebleeding was noted in 40% (2/5). The 30-day mortality in this group was 20% (1/5). The remaining 58% (7/12) had hemostatic spray in conjunction with the combined therapy (DH). Of these, the initial hemostasis rate was 86% (6/7). Rebleeding occurred in 17% (1/6). There was no mortality in this group.

      Forrest 1b

      Thirty-eight percent (3/8) of patients with Forrest 1b lesions had AH therapy, with initial hemostasis of 100%, 0% rebleeding, and 0% 30-day mortality (Table 2). The remaining 62% (5/8) received DH therapy. Initial hemostasis was 100%, and there were no rebleeding episodes. However, 30-day mortality was 40% (2/5). One death was due to septicemia and another was due to aspiration pneumonia.

      Other adverse events and technical failure with hemostatic spray

      There was no significant hemostatic spray-related equipment failure. On occasion, it was necessary to use the second catheter supplied to achieve adequate therapy. This was due to blockage within the first catheter induced by moisture within the endoscope channel or the gut lumen. The former can be minimized by vigorous flushing of the endoscope channel with air before catheter insertion. No other adverse event was observed.

      Conventional therapy group outcomes (August 2010 to June 2013)

      Baseline characteristics (age, gender, cause, preceding antithrombotic use, Rockall and Blatchford scores) were similar in both groups (Table 3). Descriptions of individual cases are provided in Supplementary Table 2, available online at www.giejournal.org.
      Table 3Clinical characteristics of patients with Forrest 1a and 1b ulcers in the hemostatic spray and conventional therapy groups
      Clinical characteristicsHemostatic spray group (July 2013 to August 2015) (n = 20)Conventional therapy (August 2010 to June 2013) (n = 20)P value
      Age (years)75 (11)73 (21).72
      Women1091.00
      Rockall score8 (2)7 (3).30
      Blatchford score14 (6)14 (4).82
      Use of antithrombotic therapy1012.75
      Data presented as median (interquartile range) or number.

      Initial hemostasis and rebleeding

      Initial hemostasis was achieved with conventional strategies in 16 of 20 patients (80%) and rebleeding occurred in 4 of 16 patients (25%) and 5 of 16 patients (31%) at 7 days and 30 days, respectively. Three of 4 patients with failed initial hemostasis required emergency laparotomy. One died of ongoing bleeding. Three patients were treated with adrenaline monotherapy due to technical difficulties in deploying clips or a thermal device. Rebleeding occurred in 2 of 3 (67%) of these cases. Among those who received the standard dual therapy (adrenaline and hemostatic clips [AC] or adrenaline and a thermal device [AT]), initial hemostasis was achieved in 14 of 17 patients (82%). The 30-day rebleeding rate of 21% (3/14) was comparatively higher than that seen with the combination of hemostatic spray and standard dual therapy (9%).

      Mortality

      The overall 30-day mortality with conventional therapy was 30% (6/20), and GI bleed-related death occurred in 15% (3/20).

      Discussion

      Our study examined the efficacy of Hemospray as a second agent to adrenaline and as a third agent to combined therapy in high-risk patients with NVUGIB. These patients were truly high risk with median Rockall and Blatchford scores of 8 and 14, respectively. In this context, our rationale was to use hemostatic spray to maximize the hemostatic effect. Our data highlight 2 important points. First, using this approach, hemostatic spray was effective in attaining initial hemostasis as shown by the favorable rate of 95%. Second, hemostatic spray offers good protection against rebleeding. Indeed, the overall rebleeding rate in our cohort was 16%, and when used in combination with dual conventional treatment modalities, the rebleeding rate was 9%.
      In this study, hemostatic spray was used as a second agent to adrenaline in 40% of cases in which the deployment of clips or thermal contact devices was not technically feasible. Initial hemostasis achieved in combination with adrenaline was 100%, even in high-risk lesions. Importantly, the rebleeding rate of 25% in Forrest 1a and Forrest 1b ulcers with this combination was better than that seen with adrenaline monotherapy for similar lesions.
      • Lin H.J.
      • Hsieh Y.H.
      • Tseng G.Y.
      • et al.
      Endoscopic injection with fibrin sealant versus epinephrine for arrest of peptic ulcer bleeding: a randomized, comparative trial.
      • Pescatore P.
      • Jornod P.
      • Borovicka J.
      • et al.
      Epinephrine versus epinephrine plus fibrin glue injection in peptic ulcer bleeding: a prospective randomized trial.
      When compared with our previous experience in a group of patients with Forrest 1a and 1b lesions before the introduction of hemostatic spray in our unit, rebleeding occurred in 67% when the therapeutic option was technically limited to adrenaline monotherapy. Therefore, hemostatic spray may be a viable alternative for lesions inaccessible to hemostatic methods that require precise targeting. To this end, hemostatic spray can be applied without direct mucosal contact or en face view of a lesion. Prior injection of adrenaline can provide immediate clear visualization for a targeted application of hemostatic spray and complement its hemostatic effect. Thus, hemostatic spray offers ease of application, particularly in technically challenging cases. It obviates the need for more advanced techniques such as the use of a duodenoscope or a cap-assisted approach, for which the required skills may not be widely available.
      Sustained hemostasis with a combination of hemostatic spray and adrenaline in our cohort was comparatively lower than that reported with conventional measures, particularly for Forrest 1a ulcers.
      • Chung I.-K.
      • Ham J.-S.
      • Kim H.-S.
      • et al.
      Comparison of the hemostatic efficacy of the endoscopic hemoclip method with hypertonic saline–epinephrine injection and a combination of the two for the management of bleeding peptic ulcers.
      • Giday S.A.
      • Kim Y.
      • Krishnamurty D.M.
      • et al.
      Long-term randomized controlled trial of a novel nanopowder hemostatic agent (TC-325) for control of severe arterial upper gastrointestinal bleeding in a porcine model.
      • Chung S.S.
      • Lau J.Y.
      • Sung J.J.
      • et al.
      Randomised comparison between adrenaline injection alone and adrenaline injection plus heat probe treatment for actively bleeding ulcers.
      • Lin H.J.
      • Tseng G.Y.
      • Perng C.L.
      • et al.
      Comparison of adrenaline injection and bipolar electrocoagulation for the arrest of peptic ulcer bleeding.
      • Park C.H.
      • Joo Y.E.
      • Kim H.S.
      • et al.
      A prospective, randomized trial comparing mechanical methods of hemostasis plus epinephrine injection to epinephrine injection alone for bleeding peptic ulcer.
      Although Hemospray has been licensed as monotherapy for NVUGIB, its efficacy in Forrest 1a lesions has not been adequately studied. Our experience of hemostatic spray as monotherapy was limited to a single case where it was successfully used to control a significant bleeding episode from pangastritis. In a porcine model, hemostatic spray induces 100% hemostasis with 20% rebleeding within 24 hours.
      • Giday S.A.
      • Kim Y.
      • Krishnamurty D.M.
      • et al.
      Long-term randomized controlled trial of a novel nanopowder hemostatic agent (TC-325) for control of severe arterial upper gastrointestinal bleeding in a porcine model.
      These findings were corroborated by subsequent clinical studies. For example, Sung et al
      • Sung J.J.
      • Luo D.
      • Wu J.C.
      • et al.
      Early clinical experience of the safety and effectiveness of Hemospray in achieving hemostasis in patients with acute peptic ulcer bleeding.
      achieved 95% initial hemostasis with 85% sustained hemostasis at 72 hours in patients with Forrest 1a and 1b ulcers. Similarly, Smith et al
      • Smith L.A.
      • Stanley A.J.
      • Bergman J.J.
      • et al.
      Hemospray application in nonvariceal upper gastrointestinal bleeding: results of the survey to evaluate the application of Hemospray in the luminal tract.
      reported initial hemostasis in 85% and a rebleeding rate of 15% using hemostatic spray as monotherapy. However, in a subsequent case series, the rate of rebleeding with hemostatic spray monotherapy was considerable prompting authors to recommend either a second re-look or prolonged hospital observation.
      • Dixon S.
      • Tate D.
      • Przemioslo R.
      • et al.
      PTH-034 Hemospray may not reliably achieve hemostasis beyond 48 hours in acute upper gastrointestinal bleeding.
      This raises the question as to whether hemostatic spray as monotherapy is a cost-effective approach. This was addressed in a recent study, which showed that the combination of hemostatic spray with conventional hemostatic measures was more cost-effective than hemostatic spray as monotherapy.
      • Barkun A.N.
      • Adam V.
      • Chen Y.-I.
      • et al.
      Tu1736 The cost-effectiveness of hemospraytm in patients with non variceal upper gastrointestinal bleeding.
      In a recent case series, the use of hemostatic spray as monotherapy in Forrest 1a lesions was associated with 100% rebleeding, necessitating further endotherapy or interventional radiology.
      • Nasr I.
      • De Martino S.
      • Borrow D.-M.
      • et al.
      PWE-007 When to use hemospray? a single centre experience.
      Therefore, the available evidence suggests that hemostatic spray monotherapy is unlikely to replace conventional hemostatic therapy.
      In our series, hemostatic spray was used as an adjunct to the standard dual therapy in cases where the risk of rebleeding was deemed to be significant. This triple therapy approach resulted in overall initial hemostasis of 92% with a rebleeding rate of 9%. Even among those with Forrest 1a ulcers, the initial hemostasis and the rebleeding rate were encouraging. Furthermore, in Forrest 1a and 1b lesions, the rebleeding rate of 9% was comparatively less than the rebleeding rate reported with the conventional approaches for similar lesions.
      • Masci E.
      • Arena M.
      • Morandi E.
      • et al.
      Upper gastrointestinal active bleeding ulcers: review of literature on the results of endoscopic techniques and our experience with Hemospray.
      • Pescatore P.
      • Jornod P.
      • Borovicka J.
      • et al.
      Epinephrine versus epinephrine plus fibrin glue injection in peptic ulcer bleeding: a prospective randomized trial.
      • Chau C.H.
      • Siu W.T.
      • Law B.K.
      • et al.
      Randomized controlled trial comparing epinephrine injection plus heat probe coagulation versus epinephrine injection plus argon plasma coagulation for bleeding peptic ulcers.
      Indeed, the hemostatic effect with the triple therapy surpassed the initial hemostasis of 82% and the rebleeding rate of 21% previously observed with conventional dual therapy in our institution. This effect might be explained by the ability of hemostatic spray to cover large areas and address multiple bleeding points from feeding vessels simultaneously. It could be surmised that the spray provides a sealant barrier that complements tamponade, vasoconstriction, and mechanical or thermal coagulation resulting from conventional combination therapy. Video 1 illustrates a typical case.
      Similar to the findings of a recent UK-wide audit, the patients in our cohort were elderly (median age of 75 years) and the mortality observed was largely related to decompensation of comorbid conditions.
      • Hearnshaw S.A.
      • Logan R.F.
      • Lowe D.
      • et al.
      Acute upper gastrointestinal bleeding in the UK: patient characteristics, diagnoses and outcomes in the 2007 UK audit.
      One GI bleed-related death was attributable to rebleeding of a duodenal ulcer. In this case, no further intervention was deemed appropriate due to significant frailty. In our study, there was no difference in mortality when hemostatic spray was used as an adjunct to adrenaline or incorporated into the standard combination therapy. Even though there have been reports of perforated viscus, biliary orifice obstruction, and splenic infarct associated with the use of hemostatic spray, we did not observe any adverse events in our cohort.
      • Yau A.H.
      • Ou G.
      • Galorport C.
      • et al.
      Safety and efficacy of Hemospray(R) in upper gastrointestinal bleeding.
      • Smith L.A.
      • Morris J.
      • Stanley A.J.
      PWE-054 Preliminary experience of hemospray in the management of diffuse portal hypertensive bleeding.
      This study has several limitations. The retrospective nature precludes a comparison of the efficacy of hemostatic spray as monotherapy compared with its use as an adjuvant to dual therapy. Cases were highly selected and treatment was at the discretion of the endoscopist rather than directed by a specific protocol. Data regarding the exact amount of hemostatic spray applied are unavailable. Our case series, albeit small, is the largest to date that reports the complementary role of hemostatic spray, either as a second agent after adrenaline injection or as a third agent after conventional dual therapy in high-risk NVUGIB.
      In conclusion, our data support the use of hemostatic spray in high-risk peptic ulcer bleeding, particularly when the application of additional hemostatic agents after adrenaline injection may not be technically feasible. Its use as an adjunct can be considered when primary hemostasis is not achieved with standard combination therapy. Whether it can replace conventional measures remains to be seen, and its role in additionally securing hemostasis in high-risk patients remains unclear. Future prospective randomized controlled trials comparing hemostatic spray alone and in various combinations with conventional therapies in high-risk bleeding lesions are required to establish its optimal place in the hemostatic armamentarium.

      Appendix

      Supplementary Table 1Endoscopic treatment and outcomes of 20 consecutive patients with Forrest 1a and 1b ulcers treated with hemostatic spray from July 2013 to August 2015
      No.Age (years)SexRockall scoreBlatchford scoreATTCauseForrest classificationEndoscopic treatmentInitial hemostasisRebleedingOutcome
      181M107YesDuodenal ulcer1bDHYesNoSurvived 30 days
      270F916NoDuodenal ulcer1aDHNo; embolizedNoSurvived 30 days
      384M1016YesDuodenal ulcer1aDHYesNoSurvived 30 days
      487F98YesDual pathology1bDHYesNoNon-GI-bleed-related death
      555M510NoDuodenal ulcer1aDHYesNoSurvived 30 days
      686F919NoDuodenal ulcer1aAHYesDay 0; too frail for treatmentGI bleed death
      771F815YesDuodenal ulcer1aDHYesNoSurvived 30 days
      875F69NoEsophageal ulcer1aAHYesNoSurvived 30 days
      952F815NoEsophageal ulcer1bDHYesNoNon-GI-bleed-related death
      1068F617NoDuodenal ulcer1aDHYesYes; day 6 embolizationSurvived 30 days
      1164M89YesEsophageal ulcer1bAHYesNoSurvived 30 days
      1274M814YesDuodenal ulcer1bAHYesNoSurvived 30 days
      1356M513YesDuodenal ulcer1aAHYesNoSurvived 30 days
      1476F615YesDuodenal ulcer1aAHYesYes; day 0 embolizationSurvived 30 days
      1576F617NoGastric ulcer1bDHYesNoSurvived 30 days
      1627M410NoDuodenal ulcer1aDHYesNoSurvived 30 days
      1776M812YesEsophageal ulcer1bAHYesNoSurvived 30 days
      1875M814NoAmpullectomy1aAHYesNoSurvived 30 days
      1979M617NoMallory-Weiss1bDHYesNoSurvived 30 days
      2066F814YesDuodenal ulcer1aDHYesNoSurvived 30 days
      ATT, Antithrombotic therapy; M, male; AH, adrenaline plus hemostatic spray; F, female; DH, conventional dual therapy plus hemostatic spray.
      Supplementary Table 2Endoscopic treatment and outcomes of 20 consecutive patients with Forrest 1a and 1b ulcers from August 2010 to June 2013 before the introduction of hemostatic spray
      No.Age (years)SexRockall scoreBlatchford scoreATTCauseForrest classificationEndoscopic treatmentInitial hemostasisRebleedingOutcome
      173F916YesDuodenal ulcer1bAYesDay 1; treated endoscopicallyGI bleed death
      268M614YesDuodenal ulcer1bATYesDay 6; treated endoscopicallySurvived 30 days
      384M712YesDuodenal ulcer1bACYesNoNon-GI bleed death
      455M512NoDuodenal ulcer1aACNo; proceeded to laparotomyNoSurvived 30 days
      586F920YesDuodenal ulcer1bATYesNoSurvived 30 days
      671M612YesDuodenal ulcer1bAYesDay 1; treated endoscopicallyNon-GI bleed death
      791F717NoDuodenal ulcer1bANoNoGI bleed death
      852M413YesDuodenal ulcer1bATYesNoSurvived 30 days
      986F915YesGastric ulcer1bATYesNoSurvived 30 days
      1058F59NoDuodenal ulcer1aATNo; proceeded to laparotomyNoSurvived 30 days
      1174M610YesDuodenal ulcer1aATYesNoSurvived 30 days
      1264F812NoSphincterotomy1aACYesNoSurvived 30 days
      1372F414NoDuodenal ulcer1aACYesNoSurvived 30 days
      1450M311NoDuodenal ulcer1aATYesNoSurvived 30 days
      1579F917NoDuodenal ulcer1aACYesDay 8; too frail for treatmentGI bleed death
      1671F613NoDuodenal ulcer1aATNo; proceeded to laparotomyNoNon-GI bleed death
      1789M916YesDuodenal ulcer1aATYesNoSurvived 30 days
      1862M816YesDuodenal ulcer1aATYesNoSurvived 30 days
      1991M714YesDuodenal ulcer1aATYesNoSurvived 30 days
      2077M615YesDuodenal ulcer1aATYesDay 2; treated endoscopicallySurvived 30 days
      ATT, Antithrombotic therapy; F, female; A, adrenaline only; M, male; AT, adrenaline plus a thermal device; AC, adrenaline plus hemostatic clip.
      Supplementary Table 3Previous studies on hemostatic spray as a monotherapy in high-risk nonvariceal upper gastrointestinal bleed
      No. of patientsLesionInitial hemostasis (%)7-day rebleeding (%)30-day mortality (%)Reference
      10Spurting vessel
      In animal (pig) model.
      1000NA
      • Giday S.A.
      • Kim Y.
      • Krishnamurty D.M.
      • et al.
      Long-term randomized controlled trial of a novel nanopowder hemostatic agent (TC-325) for control of severe arterial upper gastrointestinal bleeding in a porcine model.
      20Forrest 1a and 1b95110
      • Sung J.J.
      • Luo D.
      • Wu J.C.
      • et al.
      Early clinical experience of the safety and effectiveness of Hemospray in achieving hemostasis in patients with acute peptic ulcer bleeding.
      7Forrest 1a and 1b10014NA
      • Masci E.
      • Arena M.
      • Morandi E.
      • et al.
      Upper gastrointestinal active bleeding ulcers: review of literature on the results of endoscopic techniques and our experience with Hemospray.
      16Forrest 1a and 1b19NA
      • Smith L.A.
      • Stanley A.J.
      • Bergman J.J.
      • et al.
      Hemospray application in nonvariceal upper gastrointestinal bleeding: results of the survey to evaluate the application of Hemospray in the luminal tract.
      12Forrest 1a and 1b5829
      Unsustained primary hemostasis.
      NA
      • Nasr I.
      • De Martino S.
      • Borrow D.-M.
      • et al.
      PWE-007 When to use hemospray? a single centre experience.
      NA, Not available.
      In animal (pig) model.
      Unsustained primary hemostasis.

      Supplementary data

      References

        • Rollhauser C.
        • Fleischer D.E.
        Current status of endoscopic therapy for ulcer bleeding.
        Baillieres Best Pract Res Clin Gastroenterol. 2000; 14: 391-410
        • Hearnshaw S.A.
        • Logan R.F.
        • Lowe D.
        • et al.
        Acute upper gastrointestinal bleeding in the UK: patient characteristics, diagnoses and outcomes in the 2007 UK audit.
        Gut. 2011; 60: 1327-1335
        • Rockall T.A.
        • Logan R.F.
        • Devlin H.B.
        • et al.
        Incidence of and mortality from acute upper gastrointestinal haemorrhage in the United Kingdom. Steering Committee and members of the National Audit of Acute Upper Gastrointestinal Haemorrhage.
        BMJ. 1995; 311: 222-226
        • Marmo R.
        • Rotondano G.
        • Piscopo R.
        • et al.
        Dual therapy versus monotherapy in the endoscopic treatment of high-risk bleeding ulcers: a meta-analysis of controlled trials.
        Am J Gastroenterol. 2007; 102 (quiz 469): 279-289
        • Sung J.J.
        • Tsoi K.K.
        • Lai L.H.
        • et al.
        Endoscopic clipping versus injection and thermo-coagulation in the treatment of non-variceal upper gastrointestinal bleeding: a meta-analysis.
        Gut. 2007; 56: 1364-1373
        • Gralnek I.M.
        • Dumonceau J.M.
        • Kuipers E.J.
        • et al.
        Diagnosis and management of nonvariceal upper gastrointestinal hemorrhage: European Society of Gastrointestinal Endoscopy (ESGE) Guideline.
        Endoscopy. 2015; 47: a1-a46
        • Vergara M.
        • Bennett C.
        • Calvet X.
        • et al.
        Epinephrine injection versus epinephrine injection and a second endoscopic method in high-risk bleeding ulcers.
        Cochrane Database Syst Rev. 2014; 10: CD005584
        • Barkun A.N.
        • Moosavi S.
        • Martel M.
        Topical hemostatic agents: a systematic review with particular emphasis on endoscopic application in GI bleeding.
        Gastrointest Endosc. 2013; 77: 692-700
        • Giday S.A.
        Preliminary data on the nanopowder hemostatic agent TC-325 to control gastrointestinal bleeding.
        Gastroenterol Hepatol (N Y). 2011; 7: 620-622
        • Holster I.L.
        • van Beusekom H.M.
        • Kuipers E.J.
        • et al.
        Effects of a hemostatic powder hemospray on coagulation and clot formation.
        Endoscopy. 2015; 47: 638-645
        • Smith L.A.
        • Stanley A.J.
        • Bergman J.J.
        • et al.
        Hemospray application in nonvariceal upper gastrointestinal bleeding: results of the survey to evaluate the application of Hemospray in the luminal tract.
        J Clin Gastroenterol. 2014; 48: e89-e92
        • Sung J.J.
        • Luo D.
        • Wu J.C.
        • et al.
        Early clinical experience of the safety and effectiveness of Hemospray in achieving hemostasis in patients with acute peptic ulcer bleeding.
        Endoscopy. 2011; 43: 291-295
        • Masci E.
        • Arena M.
        • Morandi E.
        • et al.
        Upper gastrointestinal active bleeding ulcers: review of literature on the results of endoscopic techniques and our experience with Hemospray.
        Scand J Gastroenterol. 2014; 49: 1290-1295
        • Yau A.H.
        • Ou G.
        • Galorport C.
        • et al.
        Safety and efficacy of Hemospray(R) in upper gastrointestinal bleeding.
        Can J Gastroenterol Hepatol. 2014; 28: 72-76
        • Holster I.L.
        • Kuipers E.J.
        • Tjwa E.T.
        Hemospray in the treatment of upper gastrointestinal hemorrhage in patients on antithrombotic therapy.
        Endoscopy. 2013; 45: 63-66
        • Lin H.J.
        • Hsieh Y.H.
        • Tseng G.Y.
        • et al.
        Endoscopic injection with fibrin sealant versus epinephrine for arrest of peptic ulcer bleeding: a randomized, comparative trial.
        J Clin Gastroenterol. 2002; 35: 218-221
        • Pescatore P.
        • Jornod P.
        • Borovicka J.
        • et al.
        Epinephrine versus epinephrine plus fibrin glue injection in peptic ulcer bleeding: a prospective randomized trial.
        Gastrointest Endosc. 2002; 55: 348-353
        • Chung I.-K.
        • Ham J.-S.
        • Kim H.-S.
        • et al.
        Comparison of the hemostatic efficacy of the endoscopic hemoclip method with hypertonic saline–epinephrine injection and a combination of the two for the management of bleeding peptic ulcers.
        Gastrointest Endosc. 1999; 49: 13-18
        • Giday S.A.
        • Kim Y.
        • Krishnamurty D.M.
        • et al.
        Long-term randomized controlled trial of a novel nanopowder hemostatic agent (TC-325) for control of severe arterial upper gastrointestinal bleeding in a porcine model.
        Endoscopy. 2011; 43: 296-299
        • Chung S.S.
        • Lau J.Y.
        • Sung J.J.
        • et al.
        Randomised comparison between adrenaline injection alone and adrenaline injection plus heat probe treatment for actively bleeding ulcers.
        BMJ. 1997; 314: 1307-1311
        • Lin H.J.
        • Tseng G.Y.
        • Perng C.L.
        • et al.
        Comparison of adrenaline injection and bipolar electrocoagulation for the arrest of peptic ulcer bleeding.
        Gut. 1999; 44: 715-719
        • Park C.H.
        • Joo Y.E.
        • Kim H.S.
        • et al.
        A prospective, randomized trial comparing mechanical methods of hemostasis plus epinephrine injection to epinephrine injection alone for bleeding peptic ulcer.
        Gastrointest Endosc. 2004; 60: 173-179
        • Dixon S.
        • Tate D.
        • Przemioslo R.
        • et al.
        PTH-034 Hemospray may not reliably achieve hemostasis beyond 48 hours in acute upper gastrointestinal bleeding.
        Gut. 2015; 64: A420-A421
        • Barkun A.N.
        • Adam V.
        • Chen Y.-I.
        • et al.
        Tu1736 The cost-effectiveness of hemospraytm in patients with non variceal upper gastrointestinal bleeding.
        Gastrointest Endosc. 2015; 81 ([abstract]): AB576-AB577
        • Nasr I.
        • De Martino S.
        • Borrow D.-M.
        • et al.
        PWE-007 When to use hemospray? a single centre experience.
        Gut. 2015; 64: A212-A213
        • Chau C.H.
        • Siu W.T.
        • Law B.K.
        • et al.
        Randomized controlled trial comparing epinephrine injection plus heat probe coagulation versus epinephrine injection plus argon plasma coagulation for bleeding peptic ulcers.
        Gastrointest Endosc. 2003; 57: 455-461
        • Smith L.A.
        • Morris J.
        • Stanley A.J.
        PWE-054 Preliminary experience of hemospray in the management of diffuse portal hypertensive bleeding.
        Gut. 2013; 62: A152-A153
        • Forrest J.A.
        • Finlayson N.D.
        • Shearman D.J.
        Endoscopy in gastrointestinal bleeding.
        Lancet. 1974; 2: 394-397
        • Peter S.
        • Wilcox C.M.
        Modern endoscopic therapy of peptic ulcer bleeding.
        Dig Dis. 2008; 26: 291-299
        • Laine L.
        • Peterson W.L.
        Bleeding peptic ulcer.
        N Engl J Med. 1994; 331: 717-727